I agree with all that, of course.
Ecologically, with a database large/robust enough, algorithms could be developed to discover likelihood of host or pollinator associations, particularly for ephemeral species based on timing of photo/data of arthropod and host, especially where there is a range of observability over geographic space.
Abundance, this metric/index would be difficult as the field is not included in the Observation Submission. However, Ebird uses an interesting measure for frequency for their phenology/bar charts. I think a lot of people do not realize this, but their bar charts are based on the percentage of checklists which the said species appears on, regardless of 10 or 10 thousand individuals. Perhaps a good iNat adaption to this would be percentage of times a said taxon is documented within a participating member's observations during a 24 hour period.
The stat visualization that I was originally talking about was to deal more with determining the effect of user interaction on organized biogeographical information. Also to measure the effect that iNat is having on its users; which I would assume the site is making many of us better biogeographers.